Reliable Moderate HV testing

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Jonathan Peakall by way of Terry Fritz <twftesla-at-qwest-dot-net>" <jpeakall-at-madlabs.info>

Hi All,
 
I am having trouble measuring the output of my flyback system. The measurements
vary widely. I am using a 100:1 HV probe marked " X100 For 10mohm inputs model
1MA-100-10" and a Metex M-4650CR DMM, rated to 200kHz. When I hold the probe, I
get much lower readings when measuring AC off the flyback. I understand that I
am capacitively coupled to the probe and grounding it, but which reading is
"real" me hanging on to it or not? Which represents the "working" voltage of
the system?
 
Also, while hanging on to the probe, if I put the flyback output through a full
wave bridge, I get much less than 1/2 of what I read on the AC side.
 
Further, when reading the AC output of the system, if I disconnect the ground
lead, sparks will jump a good distance from the lead to the ground. This makes
me nervous. I tried using my mains ground, which should be isolated from the RF
ground of the system, and the same thing happens. What if I put a HV probe on
the ground side? Is my DMM at serious risk?
 
So what can I do to get at least consistent if not accurate readings?
 
Thanks all,
 
Jonathan Peakall